BirdNET confirmed some of my limited knowledge of birds. I could readily identify the chaffinch and was able to recognize the amazing variations in the sounds of some other common species, including the European robin. Occasionally, the app identified some species unknown to me, such as the Eurasian tree creeper. It would identify birds I barely heard and could not even see. And, once, it thrilled me by reporting that the sound I had just recorded was that of a peregrine falcon — the fastest diving animal on Earth. Wow! Helpfully, the app then added that this identification was “highly uncertain”, more of a “wild guess” (yes, I am sure, given my town-centre location).

The app was far from perfect. It often could not even hazard a guess because of poor recording quality or limitations in the online database that the app accessed to try to identify species. But not yet being particularly good was also part of the point, as I later learned. The state of scientific knowledge of bird sounds was, until recently, based on a relatively limited set of recordings, assembled by a small number of experts devoting their time to researching different species. BirdNET was an early effort to change this by bringing modern technology and tens of thousands of humans together into a project to collectively gather recordings and develop better ways to analyse them.

BirdNET 证实了我对鸟类的一些有限了解。我很容易识别出燕雀，并且能够识别出其他一些常见物种（包括欧洲知更鸟）的声音的惊人变化。偶尔，该应用程序会识别出一些我不知道的物种，例如欧亚爬树虎。它会识别我几乎听不到甚至看不到的鸟。有一次，我很高兴地报告说，我刚刚录制的声音是游隼的声音——地球上最快的潜水动物。哇！有用的是，该应用程序随后补充说，这种身份是 “高度不确定的”，更像是一个 “疯狂的猜测”（是的，我敢肯定，考虑到我位于市中心）。

该应用程序远非完美。由于记录质量差或应用程序访问的在线数据库存在限制，因此它通常甚至无法冒险猜测。但正如我后来了解到的那样，还不是特别好也是重点的一部分。直到最近，鸟类声音的科学知识状况还是基于一组相对有限的录音，这些录音是由少数致力于研究不同物种的专家收集的。BirdNET 是改变这种情况的早期努力，它将现代技术和数以万计的人聚集在一起，共同收集记录并开发更好的方法来分析它们。